Minimum Wages and Employment: A Case Study of the Fast-Food Industry in New Jersey and Pennsylvania
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Minimum Wages and Employment:
A Case Study of the Fast-Food Industry
in New Jersey and Pennsylvania
On April 1, 1992, New Jersey's minimum wage rose from $4.25 to $5.05 per
hour. To evaluate the impact of the law we surveyed 410 fast-food restaurants in
New Jersey and eastern Pennsylvania before and after the rise. Comparisons of
employment growth at stores in New Jersey and Pennsylvania (where the
minimum wage was constant) provide simple estimates of the effect of the higher
minimum wage. We also compare employment changes at stores in New Jersey
that were initially paying high wages (above $5) to the changes at lower-wage
stores. We find no indication that the rise in the minimum wage reduced
employment. (JEL 530, 523)
How do employers in a low-wage labor cent studies that rely on a similar compara-
market respond to an increase in the mini- tive methodology have failed to detect a
mum wage? The prediction from conven- negative employment effect of higher mini-
tional economic theory is unambiguous: a mum wages. Analyses of the 1990-1991 in-
rise in the minimum wage leads perfectly creases in the federal minimum wage
competitive employers to cut employment (Lawrence F. Katz and Krueger, 1992; Card,
(George J. Stigler, 1946). Although studies 1992a) and of an earlier increase in the
in the 1970's based on aggregate teenage minimum wage in California (Card, 1992b)
employment rates usually confirmed this find no adverse employment impact. A study
prediction,' earlier studies based on com- of minimum-wage floors in Britain (Stephen
parisons of employment at affected and un- Machin and Alan Manning, 1994) reaches a
affected establishments often did not (e.g., similar conclusion.
Richard A. Lester, 1960, 1964). Several re- This paper presents new evidence on the
effect of minimum wages on establishment-
level employment outcomes. We analyze the
experiences of 410 fast-food restaurants in
*Department of Economics, Princeton University, New Jersey and Pennsylvania following the
Princeton, NJ 08544. We are grateful to the Institute increase in New Jersey's minimum wage
for Research on Poverty, University of Wisconsin, for from $4.25 to $5.05 per hour. Comparisons
partial financial support. Thanks to Orley Ashenfelter,
Charles Brown, Richard Lester, Gary Solon, two
of employment, wages, and prices at stores
anonymous referees, and seminar participants at in New Jersey and Pennsylvania before and
Princeton, Michigan State, Texas A&M, University of after the rise offer a simple method for
Michigan, university of Pennsylvania, ~niversitJ of evaluating the effects of the-minimum wage.
Chicago, and the NBER for comments and sugges- ~~~~~~i~~~~ within N~~ jerseybetween
tions. We also acknowledge the expert research assis-
tance of Susan Belden, Chris Burris, Geraldine Harris, high-wage paying
and Jonathan Orszag. than the new minimum rate prior to its
'see Charles Brown et al. (1982,1983) for surveys of effective date) and other stores provide an
this literature. A recent update (Allison J. Wellington, alternative estimate of the impact of the
1991) concludes that the employment effects of the
minimum wage are negative but small: a 10-percent
new lawe
increase in the minimum is estimated to lower teenage In addition to the simplicity of our empir-
employment rates by 0.06 percentage points. ical methodology, several other features of
772VOL. 84 NO. 4 CARD AND KRUEGER: MINIMUM WAGE AND EMPLOYMENT 773
the New Jersey law and our data set are and 1991 to measures of the relative mini-
also significant. First, the rise in the mini- mum wage in each state.
mum wage occurred during a recession. The
increase had been legislated two years ear- I. The New Jersey Law
lier when the state economy was relatively
healthy. By the time of the actual increase, A bill signed into law in November 1989
the unemployment rate in New Jersey had raised the federal minimum wage from $3.35
risen substantially and last-minute political per hour to $3.80 effective April 1, 1990,
action almost succeeded in reducing the with a further increase to $4.25 per hour on
minimum-wage increase. It is unlikely that April 1, 1991. In early 1990 the New Jersey
the effects of the higher minimum wage legislature went one step further, enacting
were obscured by a rising tide of general parallel increases in the state minimum wage
economic conditions. for 1990 and 1991 and an increase to $5.05
Second, New Jersey is a relatively small per hour effective April 1, 1992. The sched-
state with an economy that is closely linked uled 1992 increase gave New Jersey the
to nearby states. We believe that a control highest state minimum wage in the country
group of fast-food stores in eastern Pennsyl- and was strongly opposed by business lead-
vania forms a natural basis for comparison ers in the state (see Bureau of National
with the experiences of restaurants in New Affairs, Daily Labor Report, 5 May 1990).
Jersey. Wage variation across stores in New In the two years between passage of the
Jersey, however, allows us to compare the $5.05 minimum wage and its effective date,
experiences of high-wage and low-wage New Jersey's economy slipped into reces-
stores within New Jersey and to test the sion. Concerned with the potentially ad-
validity of the Pennsylvania control group. verse impact of a higher minimum wage, the
Moreover, since seasonal patterns of em- state legislature voted in March 1992 to
ployment are similar in New Jersey and phase in the 80-cent increase over two years.
eastern Pennsylvania, as well as across The vote fell just short of the margin re-
high- and low-wage stores within New Jer- quired to override a gubernatorial veto, and
sey, our comparative methodology effec- the Governor allowed the $5.05 rate to go
tively "differences out" any. seasonal em- into effect on April 1 before vetoing the
ployment effects. two-step legislation. Faced with the prospect
Third, we successfully followed nearly 100 of having to roll back wages for minimum-
percent of stores from a first wave of inter- wage earners, the legislature dropped the
views conducted just before the rise in the issue. Despite a strong last-minute chal-
minimum wage (in February and March lenge, the $5.05 minimum rate took effect
1992) to a second wave conducted 7-8 as originally planned.
months after (in November and December
1992). We have complete information on 11. Sample Design and Evaluation
store closings and take account of employ-
ment changes at the closed stores in our Early in 1992 we decided to evaluate the
analyses. We therefore measure the overall impending increase in the New Jersey mini-
effect of the minimum wage on average mum wage by surveying fast-food restau-
employment, and not simply its effect on rants in New Jersey and eastern Pennsylva-
surviving establishments. niae2 Our choice of the fast-food industry
-Our analysis of employment trends at was driven by several factors. First, fast-food
stores that were open for business before stores are a leading employer of low-wage
the increase in the minimum wage ignores workers: in 1987, franchised restaurants em-
any potential effect of minimum wages on
the rate of new store openings. To assess
the likely magnitude of this effect we relate
state-specific growth rates in the number of 2At the time we were uncertain whether the $5.05
McDonald's fast-food outlets between 1986 rate would go into effect or be overridden.THE AMERICAN ECONOMIC REVIEW SEPTEMBER 1994
Stores in:
A1l NJ PA
Waue I, February 15-March 4, 1992:
Number of stores in sample frame:a 473 364 109
Number of refusals: 63 33 30
Number interviewed: 410 331 79
Response rate (percentage): 86.7 90.9 72.5
Wace 2, Nocember 5 - December 31, 1992:
Number of stores in sample frame: 410 331 79
Number closed: 6 5 1
Number under rennovation: 2 2 0
Number temporarily closed:' 2 2 0
Number of refusals: 1 1 0
Number i n t e r v i e ~ e d : ~ 399 321 78
aStores with working phone numbers only; 29 stores in original sample frame had
disconnected phone numbers.
'~ncludes one store closed because of highway construction and one store closed
because of a fire.
'Includes 371 phone interviews and 28 personal interviews of stores that refused an
initial request for a phone interview.
ployed 25 percent of all workers in the rants in New Jersey and eastern Pennsylva-
restaurant industry (see U.S. Department of nia from the Burger King, KFC, Wendy's,
Commerce, 1990 table 13). Second, fast-food and Roy Rogers chain^.^ The first wave of
restaurants comply with minimum-wage reg- the survey was conducted by telephone in
ulations and would be expected to raise late February and early March 1992, a little
wages in response to a rise in the minimum over a month before the scheduled increase
wage. Third, the job requirements and in New Jersey's minimum wage. The survey
products of fast-food restaurants are rela- included questions on employment, starting
tively homogeneous, making it easier to ob- wages, prices, and other store characteris-
tain reliable measures of employment, tic~.~
wages, and product prices. The absence of Table 1 shows that 473 stores in our sam-
tips greatly simplifies the measurement of ple frame had working telephone numbers
wages in the industry. Fourth, it is relatively when we tried to reach them in February-
easy to construct a sample frame of fran- March 1992. Restaurants were called as
chised restaurants. Finally, past experience many as nine times to elicit a response. We
(Katz and Krueger, 1992) suggested that obtained completed interviews (with some
fast-food restaurants have high response item nonresponse) from 410 of the restau-
rates to telephone survey^.^ rants, for an overall response rate of 87
Based on these considerations we con- percent. The response rate was higher in
structed a sample frame of fast-food restau- New Jersey (91 percent) than in Pennsylva-
4 ~ h seample was derived from white-pages tele-
3 ~ an pilot survey Katz and Krueger (1992) obtained phone listings for New Jersey and Pennsylvania as of
very low response rates from McDonald's restaurants. February 1992.
For this reason, McDonald's restaurants were excluded 'copies of the questionnaires used in both waves of
from Katz and Krueger's and our sample frames. the survey are available from the authors upon request.VOL. 84 NO. 4 C A m AND KRUEGER: MINIiiMUM WAGE AND EMPLOYMENT 775
nia (72.5 percent) because our interviewer nently closed stores but is treated as missing
made fewer call-backs to nonrespondents in for the temporarily closed stores. (Full-
Penn~ylvania.~ In the analysis below we in- time-equivalent [FTE] employment was cal-
vestigate possible biases associated with the culated as the number of full-time workers
degree of difficulty in obtaining the first- [including managers] plus 0.5 times the
wave interview. number of part-time workers.)' Means are
The second wave of the survey was con- presented separately for stores in New Jer-
ducted in November and December 1992, sey and Pennsylvania, along with t statistics
about eight months after the minimum-wage for the null hypothesis that the means are
increase. Only the 410 stores that re- equal in the two states.
sponded in the first wave were contacted in Rows la-e show the distribution of stores
the second round of interviews. We success- by chain and ownership status (company-
fully interviewed 371 (90 percent) of these owned versus franchisee-owned). The
stores by phone in November 1992. Because Burger King, Roy Rogers, and Wendy's
of a concern that nonresponding restaurants stores in our sample have similar average
might have closed, we hired an interviewer food prices, store hours, and employment
to drive to each of the 39 nonrespondents levels. The KFC stores are smaller and are
and determine whether the store was still open for fewer hours. They also offer a
open, and to conduct a personal interview if more expensive main course than stores in
possible. The interviewer discovered that six the other chains (chicken vs, hamburgers).
restaurants were permanently closed, two In wave 1, average employment was 23.3
were temporarily closed (one because of a full-time equivalent workers per store in
fire, one because of road construction), and Pennsylvania, compared with an average of
two were under renovation.' Of the 29 stores 20.4 in New Jersey. Starting wages were
open for business, all but one granted a very similar among stores in the two states,
request for a personal interview. As a re- although the average price of a "full meal"
sult, we have second-wave interview data (medium soda, small fries, and an entree)
for 99.8 percent of the restaurants that re- was significantly higher in New Jersey. There
sponded in the first wave of the survey, and were no significant cross-state differences in
information on closure status for 100 per- average hours of operation, the fraction of
cent of the sample. full-time workers, or the prevalence of bonus
Table 2 presents the means for several programs to recruit new worker^.^
key variables in our data set, averaged over The average starting wage at fast-food
the subset of nonmissing responses for each restaurants in New Jersey increased by 10
variable. In constructing the means, employ- percent following the rise in the minimum
ment in wave 2 is set to 0 for the perma- wage. Further insight into this change is
provided in Figure 1, which shows the dis-
tributions of starting wages in the two states
before and after the rise. In wave 1, the
distributions in New Jersey and Pennsylva-
6 ~ e s p o n s erates per call-back were almost identical nia were very similar. By wave 2 virtually all
in the two states. Among New Jersey stores, 44.5
percent responded on the first call, and 72.0 percent
responded after at most two call-backs. Among Penn-
sylvania stores 42.2 percent responded on the first call, ' w e discuss the sensitivity of our results to alterna-
and 71.6 percent responded after at most two call- tive assumptions on the measurement of employment
backs. in Section 111-C.
7 ~ ofs April 1993 the store closed because of road ' ~ h e s e programs offer current employees a cash
construction and one of the stores closed for renova- "bounty" for recruiting any new employee who stays
tion had reopened. The store closed by fire was open on the job for a minimum period of time. Typical
when our telephone interviewer called in November bounties are $50-$75. Recruiting programs that award
1992 but refused the interview. By the time of the the recruiter with an "employee of the month" desig-
follow-up personal interview a mall fire had closed the nation or other noncash bonuses are excluded from our
store. tabulations.THE AMERICAN ECONOMIC REVIEW SEPTEMBER 1994
Stores in:
Variable NJ PA ta
1. Distribution of Store Types (percentages):
a. Burger King
b. KFC
c. Roy Rogers
d. Wendy's
e. Company-owned
2. Means in Wave I:
a. FTE employment 20.4
(0.51)
b. Percentage full-time employees 32.8
(1.3)
c. Starting wage 4.61
(0.02)
d. Wage = $4.25 (percentage) 30.5
(2.5)
e. Price of full meal
f. Hours open (weekday)
g. Recruiting bonus
3. Means in Ware 2:
a. FTE employment 21.0 21.2 - 0.2
(0.52) (0.94)
b. Percentage full-time employees 35.9 30.4 1.8
(1.4) (2.8)
c. Starting wage 5.08 4.62 10.8
(0.01) (0.04)
d. Wage = $4.25 (percentage) 0.0 25.3 -
(4.9)
e. Wage = $5.05 (percentage) 85.2 1.3 36.1
(2.0) (1.3)
f. Price of full meal 3.41 3.03 5.0
(0.04) (0.07)
g. Hours open (weekday) 14.4 14.7 - 0.8
(0.2) (0.3)
h. Recruiting bonus 20.3 23.4 - 0.6
(2.3) (4.9)
Notes: See text for definitions. Standard errors are given in parentheses.
aTest of equality of means in New Jersey and Pennsylvania.
restaurants in New Jersey that had been Despite the increase in wages, full-time-
paying less than $5.05 per hour reported a equivalent employment increased in New
starting wage equal to the new rate. Inter- Jersey relative to Pennsylvania. Whereas
estingly, the minimum-wage increase had no New Jersey stores were initially smaller,
apparent "spillover" on higher-wage restau- employment gains in New Jersey coupled
rants in the state: the mean percentage wage with losses in Pennsylvania led to a small
change for these stores was - 3.1 percent. and statistically insignificant interstateVOL. 84 NO. 4 CARD AND KRUEGER: MINIMUM WAGE AND EMPLOYMENT
February 1 9 9 2
Wage Range
November 1 9 9 2
Wage Range
New Jersey Pennsylvania
FIGURE
1. DISTRIBUTION
OF STARTING
WAGERATES778 THE AMERICAN ECONOMIC REVIEW SEPTEMBER I994
difference in wave 2. Only two other vari- our survey. We present data by state in
ables show a relative change between waves columns (i) and (ii), and for stores in New
1 and 2: the fraction of full-time employees Jersey classified by whether the starting
and the price of a meal. Both variables wage in wave 1 was exactly $4.25 per hour
increased in New Jersey relative to Pennsyl- [column (iv)] between $4.26 and $4.99 per
vania. hour [column (v)] or $5.00 or more per hour
We can assess the reliability of our survey [column (vi)]. We also show the differences
questionnaire by comparing the responses in average employment between New Jersey
of 11 stores that were inadvertently inter- and Pennsylvania stores [column (iii)] and
viewed twice in the first wave of the survey.10 between stores in the various wage ranges
Assuming that measurement errors in the in New Jersey [columns (viil-(viii)].
two interviews are independent of each Row 3 of the table presents the changes
other and independent of the true variable, in average employment between waves 1
the correlation between responses gives an and 2. These entries are simply the differ-
estimate of the "reliability ratio" (the ratio ences between the averages for the two
of the variance of the signal to the com- waves (i.e., row 2 minus row 1). An alterna-
bined variance of the signal and noise). The tive estimate of the change is presented in
estimated reliability ratios are fairly high, row 4: here we have computed the change
ranging from 0.70 for full-time equivalent in employment over the subsample of stores
employment to 0.98 for the price of a meal." that reported valid employment data in both
We have also checked whether stores with waves. We refer to this group of stores as
missing data for any key variables are dif- the balanced subsample. Finally, row 5 pre-
ferent from restaurants with complete re- sents the average change in employment in
sponses. We find that stores with missing the balanced subsample, treating wave-2
data on employment, wages, or prices are employment at the four temporarily closed
similar in other respects to stores with com- stores as zero, rather than as missing.
plete data. There is a significant size differ- As noted in Table 2, New Jersey stores
ential associated with the likelihood of the were initially smaller than their Pennsylva-
store closing after wave 1. The six stores nia counterparts but grew relative to Penn-
that closed were smaller than other stores sylvania stores after the rise in the mini-
(with an average employment of only 12.4 mum wage. The relative gain (the "dif-
full-time-equivalent employees in wave 1).12 ference in differences" of the changes in
employment) is 2.76 FTE employees (or 13
111. Employment Effects of the percent), with a t statistic of 2.03. Inspec-
Minimum-Wage Increase tion of the averages in rows 4 and 5 shows
that the relative change between New Jer-
A. Differences in Differences sey and Pennsylvania stores is virtually iden-
tical when the analysis is restricted to the
Table 3 summarizes the levels and balanced subsample, and it is only slightly
changes in average employment per store in smaller when wave-2 employment at the
temporarily closed stores is treated as zero.
Within New Jersey, employment ex-
10
These restaurants were interviewed twice because panded at the low-wage stores (those paying
their phone numbers appeared in more than one phone $4.25 per hour in wave 1) and contracted at
book, and neither the interviewer nor the respondent
noticed that they were previously interviewed. the high-wage stores (those paying $5.00 or
11
Similar reliability ratios for very similar questions more per hour). Indeed, the average change
were obtained by Katz and Krueger (1992). in employment at the high-wage stores
''A probit analysis of the probability of closure ( - 2.16 FTE employees) is almost identical
shows that the initial size of the store is a significant
predictor of closure. The level of starting wages has a
to the change among Pennsylvania stores
numerically small and statistically insignificant coeffi- ( - 2.28 FTE employees). Since high-wage
cient in the probit model. stores in New Jersey should have beenV O L . 84 NO. 4 CARD AND KRUEGER: MINIMUM WAGE AND EMPLOYMENT
largely unaffected by the new minimum els of the form:
wage, this comparison provides a specifica-
tion test of the validity of the Pennsylvania (la) AE,=a+bXi+cNJi+~,
control group. The test is clearly passed.
Regardless of whether the affected stores
are compared to stores in Pennsylvania or
high-wage stores in New Jersey, the esti-
+
( l b ) AE, = a' + blXi clGAPi + E{
mated employment effect of the minimum where AE, is the change in employment
wage is similar. from wave 1 to wave 2 at store i, Xi is a set
The results in Table 3 suggest that em- of characteristics of store i, and NJ, is a
ployment contracted between February and dummy variable that equals 1 for stores in
November of 1992 at fast-food stores that New Jersey. GAP, is an alternative measure
were unaffected by the rise in the minimum of the impact of the minimum wage at store
wage (stores in Pennsylvania and stores in i based on the initial wage at that store
New Jersey paying $5.00 per hour or more (W,,):
in wave 1). We suspect that the reason for
this contraction was the continued worsen- GAP, = 0 for stores in Pennsylvania
ing of the economies of the middle-Atlantic
states during 1992.13 Unemployment rates =0 for stores in New Jersey with
in New Jersey, Pennsylvania, and New York
all trended upward between 1991 and 1993,
with a larger increase in New Jersey than
Pennsylvania during 1992. Since sales of
franchised fast-food restaurants are pro- for other stores in New Jersey.
cyclical, the rise in unemployment would be
expected to lower fast-food employment in GAP, is the proportional increase in wages
the absence of other factors.14 at store i necessary to meet the new mini-
mum rate. Variation in GAP, reflects both
B. Regression-Adjusted Models the New Jersey-Pennsylvania contrast and
differences within New Jersey based on re-
The comparisons in Table 3 make no ported starting wages in wave 1. Indeed, the
allowance for other sources of variation in value of GAP, is a strong predictor of the
employment growth, such as differences actual proportional wage change between
across chains. These are incorporated in the waves 1 and 2 (R* = 0.75), and conditional
estimates in Table 4. The entries in this on GAP, there is no difference in wage
table are regression coefficients from mod- behavior between stores in New Jersey and
Pennsylvania. l5
The estimate in column (i) of Table 4
is directly comparable to the simple
13
An alternative possibility is that seasonal factors
difference-in-differences of employment
produce higher employment at fast-food restaurants in changes in column (iv), row 4 of Table 3.
February and March than in November and December. T h e discrepancy between the two
An analysis of national employment data for food estimates is due to the restricted sample in
preparation and service workers, however, shows higher Table 4. In Table 4 and the remaining ta-
average employment in the fourth quarter than in the
first quarter. bles in this section we restrict our analysis
14
To investigate the cyclicality of fast-food restau- to the set of stores with available employ-
rant sales we regressed the year-to-year change in U.S. ment and wage data in both waves of the
sales of the McDonald's restaurant chain from
1976-1991 on the corresponding change in the unem-
ployment rate. The regression results show that a
1-percentage-point increase in the unemployment rate 1 5 regression
~ of the proportional wage change be-
reduces sales by $257 million, with a t statistic of 3.0. tween waves 1 and 2 on GAP, has a coefficient of 1.03.THE AMERICAN ECONOMIC REVlEW SEPTEMBER 1994
TABLE3-AVERAGE EMPLOYMENT PER STOREBEFOREAND I ~ E THE
R RISE
IN NEW JERSEYMINIMUM WAGE
Stores by state Stores in New Jersey a Differences within N J ~
Difference, Wage = Wage = Wage r Low- Midrange-
PA NJ NJ-PA $4.25 $4.26-$4.99 $5.00 high high
Variable (i) (ii) (iii) (iv) (v) (vi) (vii) (viii)
1. FTE employment before,
all available observations
2. FTE employment after,
all available observations
3. Change in mean FTE
employment
4. Change in mean FTE
employment, balanced
sample of storesC
5. Change in mean FTE
employment, setting
FTE at temporarily
closed stores to Od
Notes: Standard errors are shown in parentheses. The sample consists of all stores with available data on employment. FTE
(full-time-equivalent) employment counts each part-time worker as half a full-time worker. Employment at six closed stores
is set to zero. Employment at four temporarily closed stores is treated as missing.
astares in New Jersey were classified by whether starting wage in wave 1 equals $4.25 per hour ( N = 101), is between
$4.26 and $4.99 per hour ( N = 140), or is $5.00 per hour or higher ( N = 73).
b ~ i f f e r e n c ein employment between low-wage ($4.25 per hour) and high-wage ( 2$5.00 per hour) stores; and difference
in employment between midrange ($4.26-$4.99 per hour) and high-wage stores.
'Subset of stores with available employment data in wave 1 and wave 2.
this row only, wave-2 employment at four temporarily closed stores is set to 0. Employment changes are based on the
subset of stores with available employment data in wave 1 and wave 2.
TABLE4-REDUCED-FORM MODELSFOR CHANGEIN EMPLOYMENT
Model
Independent variable (i) (ii) (iii) (iv) (v)
1. New Jersey dummy 2.33 2.30 - - -
(1.19) (1.20)
2. Initial wage gapa - - 15.65 14.92 11.91
(6.08) (6.21) (7.39)
3. Controls for chain and no yes no yes yes
ownershipb
4. Controls for regionC
5. Standard error of regression
6. Probability value for controlsd
Notes: Standard errors a r e given in parentheses. T h e sample consists of 357 stores
with available data o n employment and starting wages in waves 1 and 2. T h e
dependent variable in all models is change in F T E employment. T h e mean and
standard deviation of the dependent variable are -0.237 and 8.825, respectively. All
models include a n unrestricted constant (not reported).
aProportional increase in starting wage necessary to raise starting wage t o new
minimum rate. For stores in Pennsylvania the wage gap is 0.
b ~ h r e de ummy variables for chain type and whether or not the store is company-
owned are included.
'Dummy variables for two regions of New Jersey and two regions of eastern
Pennsylvania are included.
d ~ r o b a b i l i t yv alue of joint F test for exclusion of all control variables.VOL. 84 NO. 4 CARD AND KRUEGER: MINIMUM WAGE AND EMPLOYMENT 781
survey. This restriction results in a slightly null hypothesis of a zero employment effect
smaller estimate of the relative increase in of the minimum wage. One explanation for
employment in New Jersey. this attenuation is the presence of measure-
The model in column (ii) introduces a ment error in the starting wage. Even if
set of four control variables: dummies for employment growth has no regional compo-
three of the chains and another dummy for nent, the addition of region dummies will
company-owned stores. As shown by the lead to some attenuation of the estimated
probability values in row 6, these covariates GAP coefficient if some of the true varia-
add little to the model and have no effect tion in GAP is explained by region. Indeed,
on the size of the estimated New Jersey calculations based on the estimated reliabil-
dummy. ity of the GAP variable (from the set of 11
The specifications in columns (iiil-(v) use double interviews) suggest that the fall in
the GAP variable to measure the effect of the estimated GAP coefficient from column
the minimum wage. This variable gives a (iv) to column (v) is just equal to the ex-
slightly better fit than the simple New Jer- pected change attributable to measurement
sey dummy, although its implications for the error.16
New Jersey-Pennsylvania comparison are We have also estimated the models in
similar. The mean value of GAPi among Table 4 using as a dependent variable the
New Jersey stores is 0.11. Thus the estimate proportional change in employment at each
in column (iii) implies a 1.72 increase in store.17 The estimated coefficients of the
FTE employment in New Jersey relative to New Jersey dummy and the GAP variable
Pennsylvania. are uniformly positive in these models but
Since GAP, varies within New Jersey, it is insignificantly different from 0 at conven-
possible to add both GAP, and NJ, to the tional levels. The implied employment ef-
employment model. The estimated coeffi- fects of the minimum wage are also smaller
cient of the New Jersey dummy then pro- when the dependent variable is expressed in
vides a test of the Pennsylvania control proportional terms. For example, the GAP
group. When we estimate these models, the coefficient in column (iii) of Table 4 implies
coefficient of the New Jersey dummy is in- that the increase in minimum wages raised
significant (with t ratios of 0.3-0.7), imply- employment at New Jersey stores that were
ing that inferences about the effect of the initially paying $4.25 per hour by 14 per-
minimum wage are similar whether the cent. The estimated GAP coefficient from a
comparison is made across states or across corresponding proportional model implies
stores in New Jersey with higher and lower an effect of only 7 percent. The difference is
initial wages. attributable to heterogeneity in the effect of
An even stronger test is provided in col- the minimum wage at larger and smaller
umn (v), where we have added dummies stores. Weighted versions of the propor-
representing three regions of New Jersey tional-change models (using initial employ-
(North, Central, and South) and two regions ment as a weight) give rise to wage elastici-
of eastern Pennsylvania (Allentown-Easton
and the northern suburbs of Philadelphia).
These dummies control for any region-
16
s~ecificdemand shocks and identifv the ef- In a regression model without other controls the
feet of the minimum wage by expected attenuation of the GAP coefficient due to
measurement error is the reliability ratio of GAP (yo),
employment changes at higher- and lower- which we estimate at 0.70. The expected attenuation
wage within the same region of New factor when region dummies are added to the model is
Jersey. The probability value in row 6 shows y l = (Yo- ~ 2 ) / ( 1 - ~ 2 ) where
, ~2 is the R-square
no evidence of regional components in em- statistic of a regression of GAP on region effects (equal
ployment growth. The addition of the re- to 0.30). Thus, we expect the estimated GAP coeffi-
cient to fall by a factor of Y I / Y O = 0.8 when region
gion dummies attenuates the GAP coeffi- dummies are added to a regression model.
cient and raises its standard error, however, " ~ h e s e specifications are reported in table 4 of
making it no longer possible to reject the Card and Krueger (1993).782 THE AMERICAN ECONOMIC REVIEW SEPTEMBER I994
ties similar to the elasticities implied by the little effect on the models for the level of
estimates in Table 4 (see below). employment but yield slightly smaller point
estimates in the proportional-employment-
C. Specification Tests change models.
In row 6 we present estimates obtained
The results in Tables 3 and 4 seem to from a subsample that excludes 35 stores in
contradict the standard prediction that a towns along the New Jersey shore. The ex-
rise in the minimum wage will reduce em- clusion of these stores, which may have a
ployment. Table 5 presents some alternative different seasonal pattern than other stores
specifications that probe the robustness of in our sample, leads to slightly larger mini-
this conclusion. For completeness, we re- mum-wage effects. A similar finding emerges
port estimates of models for the change in in row 7 when we add a set of dummy
employment [columns (i) and (ii)] and esti- variables that indicate the week of the
mates of models for the proportional change wave-2 inter vie^.^'
in employment [columns (iii) and (iv)].18 The As noted earlier, we made an extra effort
first row of the table reproduces the "base to obtain responses from New Jersey stores
specification" from columns (ii) and (iv) of in the first wave of our survey. The fraction
Table 4. (Note that these models include of stores called three or more times to ob-
chain dummies and a dummy for company- tain an interview was higher in New Jersey
owned stores). Row 2 presents an alterna- than in Pennsylvania. To check the sensitiv-
tive set of estimates when we set wave-2 ity of our results to this sampling feature,
employment at the temporarily closed stores we reestimated our models on a subsample
to 0 (expanding our sample size by 4). This that excludes any stores that were called
change has a small attenuating effect on the back more than twice. The results, in row 8,
coefficient of the New Jersey dummy (since are very similar to the base specification.
all four stores are in New Jersey) but less Row 9 presents weighted estimation re-
effect on the GAP coefficient (since the size sults for the proportional-employment-
of GAP is uncorrelated with the probability change models, using as weights the initial
of a temporary closure within New Jersey). levels of employment in each store. Since
Rows 3-5 present estimation results us- the proportional change in average employ-
ing alternative measures of full-time-equiv- ment is an employment-weighted average of
alent employment. In row 3, employment is the proportional changes at each store, a
redefined to exclude management employ- weighted version of the proportional-change
ees. This change has no effect relative to model should give rise to elasticities that
the base specification. In rows 4 and 5, we are similar to the implied elasticities arising
include managers in FTE employment but from the levels models. Consistent with this
reweight part-time workers as either 40 per- expectation, the weighted estimates are
cent or 60 percent of full-time workers (in- larger than the unweighted estimates, and
stead of 50 percent).19 These changes have significantly different from 0 at conventional
levels. The weighted estimate of the New
Jersey dummy (0.13) implies a 13-percent
relative increase in New Jersey employment
18
The proportional change in employment is de- -the same proportional employment effect
fined as the change in employment divided by the implied by the simple difference-in-dif-
average level of employment in waves 1 and 2. This ferences in Table 3. Similarly, the weighted
results in very similar coefficients but smaller standard
errors than the alternative of dividing by wave-1 em-
estimate of the GAP coefficient in the
ployment. For closed stores we set the proportional proportional-change model (0.81) is close to
change in employment to - 1.
19
Analysis of the 1991 Current Population Survey
reveals that part-time workers in the restaurant indus-
try work about 46 percent as many hours as full-time
20
workers. Katz and Krueger (1992) report that the ratio We also added dummies for the interview dates
of part-time workers' hours to full-time workers' hours for the wave-1 survey, but these were insignificant and
in the fast-food industry is 0.57. did not change the estimated minimum-wage effects.VOL. 84 NO. 4 CARD AND KRUEGER: MINIMUM WAGE AND EMPLOYMENT 783
Proportional change
Change in employment in employment
NJ dummy Gap measure NJ dummy Gap measure
Specification (i) (ii) (iii) (iv)
1. Base specification 2.30 14.92
(1.19) (6.21)
2. Treat four temporarily closed stores
as permanently closeda
3. Exclude managers in employment
countb
4. Weight part-time as 0.4 x full-timec
5. Weight part-time as 0.6 X full-timed
6. Exclude stores in NJ shore areae
7. Add controls for wave-2 interview
dateE
8. Exclude stores called more than twice
in wave lg
9. Weight by initial employmenth
10. Stores in towns around Newark' - 33.75
(16.75)
11. Stores in towns around CamdenJ - 10.91
(14.09)
12. Pennsylvania stores only - - 0.30
(22.00)
Notes: Standard errors are given in parentheses. Entries represent estimated coefficient of New Jersey dummy
[columns (i) and (iii)] or initial wage gap [columns (ii) and (iv)] in regression models for the change in employment
or the percentage change in employment. All models also include chain dummies and an indicator for company-
owned stores.
aWave-2 employment at four temporarily closed stores is set to 0 (rather than missing).
b~ull-timeequivalent employment excludes managers and assistant managers.
CFull-time equivalent employment equals number of managers, assistant managers, and full-time nonmanage-
ment workers, plus 0.4 times the number of part-time nonmanagement workers.
d~ull-timeequivalent employment equals number of managers, assistant managers, and full-time nonmanage-
ment workers, plus 0.6 times the number of part-time nonmanagement workers.
eSample excludes 35 stores located in towns along the New Jersey shore.
' ~ o d e l sinclude three dummy variables identifying week of wave-2 interview in November-December 1992.
gSample excludes 70 stores (69 in New Jersey) that were contacted three or more times before obtaining the
wave-1 interview.
h~egressionmodel is estimated by weighted least squares, using employment in wave 1 as a weight.
i
. Subsample of 51 stores in towns around Newark.
Subsample of 54 stores in town around Camden.
Subsample of Pennsylvania stores only. Wage gap is defined as percentage increase in starting wage necessary
to raise starting wage to $5.05.784 THE AMERICAN ECONOMIC REVIEW SEPTEMBER 1994
the implied elasticity of employment with We have also investigated whether the
respect to wages from the basic levels speci- first-differenced specification used in our
fication in row 1, column (iiI2l These find- employment models is appropriate. A
ings suggest that the proportional effect of first-differenced model implies that the level
the rise in the minimum wage was concen- of employment in period t is related to the
trated among larger stores. lagged level of employment with a coeffi-
One explanation for our finding that a cient of 1. If short-run employment fluctua-
rise in the minimum wage has a positive tions are smoothed, however, the true co-
employment effect is that unobserved de- efficient of lagged employment may be less
mand shocks within New Jersey outweighed than 1. Imposing the assumption of a unit
the negative employment effect of the mini- coefficient may then lead to biases. To test
mum wage. To address this possibility, rows the first-differenced specification we reesti-
10 and 11 present estimation results based mated models for the change in employ-
on subsamples of stores in two narrowly ment including wave-1 employment as an
defined areas: towns around Newark (row additional explanatory variable. To over-
10) and towns around Camden (row 11). In come any mechanical correlation between
each case the sample area is identified by base-period employment and the change in
the first three digits of the store's zip code.22 employment (attributable to measurement
Within both areas the change in employ- error) we instrumented wave-1 employment
ment is positively correlated with the GAP with the number of cash registers in the
variable, although in neither case is the store in wave 1 and the number of registers
effect statistically significant. To the extent in the store that were open at 11:OO A.M. In
that fast-food product market conditions are all of the specifications the coefficient of
constant within local areas, these results wave-1 employment is close to zero. For
suggest that our findings are not driven by example, in a specification including the
unobserved demand shocks. Our analysis of GAP variable and ownership and chain
price changes (reported below) also sup- dummies, the coefficient of wave-1 employ-
ports this conclusion. ment is 0.04, with a standard error of 0.24.
A final specification check is presented in We conclude that the first-differenced spec-
row 12 of Table 5. In this row we exclude ification is appropriate.
stores in New Jersey and (incorrectly) de-
fine the GAP variable for Pennsylvania D. Full-Time and Part-Time Substitution
stores as the proportional increase in wages
necessary to raise the wage to $5.05 per Our analysis so far has concentrated on
hour. In principle the size of the wage gap full-time-equivalent employment and ig-
for stores in Pennsylvania should have no nored possible changes in the distribution
systematic relation with employment growth. of full- and part-time workers. An increase
In practice, this is the case. There is no in the minimum wage could lead to an in-
indication that the wage gap is spuriously crease in full-time employment relative to
related to employment growth. part-time employment for at least two rea-
sons. First, in a conventional model one
would expect a minimum-wage increase to
induce employers to substitute skilled work-
ers and capital for minimum-wage workers.
Full-time workers in fast-food restaurants
21~ssumingaverage employment of 20.4 in New
Jersey, the 14.92 GAP coefficient in row 1, column (ii) are typically older and may well possess
im lies an employment elasticity of 0.73. higher skills than part-time workers. Thus, a
"The "070" three-digit zip-code area (around conventional model predicts that stores may
Newark) and the "080" three-digit zip-code area respond to an increase in the minimum
(around Camden) have by far the largest numbers of
stores among three-digit zip-code areas in New Jersey,
wage by increasing the proportion of full-
and together they account for 36 percent of New Jersey time workers. Nevertheless, 81 percent of
stores in our sample. restaurants paid full-time and part-timeVOL. 84 NO. 4 CARD AND KRUEGER: MINIMUM WAGE AND EMPLOYMENT 785
Regression of change in
Mean change in outcome outcome variable on:
NJ PA NJ - PA NJ dummy Wage gapa Wage gapb
Outcome measure (i) (ii) (iii) (iv) (v) (vi)
Store Characteristics:
1. Fraction full-time workersc (percentage)
2. Number of hours open per weekday
3. Number of cash registers
4. Number of cash registers open
at 11:OO A.M.
Employee Meal Programs:
5. Low-price meal program (percentage)
6. Free meal program (percentage)
7. Combination of low-price and free
meals (percentage)
Wage Profile:
8. Time to first raise (weeks)
9. Usual amount of first raise (cents)
10. Slope of wage profile (percent
per week)
Notes: Entries in columns (i) and (ii) represent mean changes in the outcome variable indicated by the row heading
for stores with available data on the outcome in waves 1 and 2. Entries in columns (iv)-(vi) represent estimated
regression coefficients of indicated variable (NJ dummy or initial wage gap) in models for the change in the
outcome variable. Regression models include chain dummies and an indicator for company-owned stores.
aThe wage gap is the proportional increase in starting wage necessary to raise the wage to the new minimum
rate. For stores in Pennsylvania, the wage gap is zero.
b ~ o d e lin
s column (vi) include dummies for two regions of New Jersey and two regions of eastern Pennsylvania.
'Fraction of part-time employees in total full-time-equivalent employment.
workers exactly the same starting wage in workers are more productive (but equally
wave 1 of our survey.23 This suggests either paid), there may be a second reason for
that full-time workers have the same skills stores to substitute full-time workers for
as part-time workers or that equity concerns part-time workers; namely, a minimum-wage
lead restaurants to pay equal wages for un- increase enables the industry to attract more
equally productive workers. If full-time full-time workers, and stores would natu-
rally want to hire a greater proportion of
full-time workers if they are more produc-
tive.
231n the other 19 percent of stores, full-time workers Row 1 of Table 6 presents the mean
are paid more, typically 10 percent more. changes in the proportion of full-time work-786 THE AMERICAN ECONOMIC REVIEW SEPTEMBER 1994
ers in New Jersey and Pennsylvania be- ployment costs unchanged. The main fringe
tween waves 1 and 2 of our survey, and benefits for fast-food employees are free
coefficient estimates from regressions of the and reduced-price meals. In the first wave
change in the proportion of full-time work- of our survey about 19 percent of fast-food
ers on the wage-gap variable, chain dum- restaurants offered workers free meals. 72
mies, a company-ownership dummy, and re- percent offered reduced-price meals, a i d 9
gion dummies [in column ( 4 1 . The results percent offered a combination of both free
are ambiguous. The fraction of full-time and reduced-price meals. Low-price meals
workers increased in New Jersey relative to are an obvious fringe benefit to cut if the
Pennsylvania by 7.3 percent (t ratio = 1.841, minimum-wage increase forces restaurants
but regressions on the wage-gap variable to pay higher wages.
show no significant shift in the fraction of Rows 5 and 6 of Table 6 present esti-
full-time workers.24 mates of the effect of the minimum-wage
increase on the incidence of free meals and
E. Other Employment-Related Measures reduced-price meals. The proportion of res-
taurants offering reduced-price meals fell
Rows 2-4 of Table 6 present results for in both New Jersey and Pennsylvania after
other outcome variables that we expect to the minimum wage increased, with a some-
be related to the level of restaurant employ- what greater decline in New Jersey. Con-
ment. In particular, we examine whether trary to an offset story, however, the reduc-
the rise in the minimum wage is associated tion in reduced-price meal programs was
with a change in the number of hours a accompanied by an increase in the fraction
restaurant is open on a weekday, the num- of stores offering free meals. Relative to
ber of cash registers in the restaurant, and stores in Pennsylvania, New Jersey employ-
the number of cash registers typically in ers actually shifted toward more generous
operation in the restaurant at 11:OO A.M. fringe benefits (i.e., free meals rather than
Consistent with our employment results, reduced-price meals). However, the relative
none of these variables shows a statistically shift is not statistically significant.
significant decline in New Jersey relative to We continue to find a statistically in-
Pennsylvania. Similarly, regressions includ- significant effect of the minimum-wage in-
ing the gap variable provide no evidence crease on the likelihood of receiving free or
that the minimum-wage increase led to a reduced-price meals in columns (v) and (vi),
systematic change in any of these variables where we report coefficient estimates of the
[see columns (v) and (vi)]. GAP variable from regression models for
the change in the incidence of these pro-
IV. Nonwage Offsets grams. The results provide no evidence that
employers offset the minimum-wage in-
One explanation of our finding that a rise crease by reducing free or reduced-price
in the minimum wage does not lower em- meals.
ployment is that restaurants can offset the Another possibility is that employers re-
effect of the minimum wage by reducing sponded to the increase in the minimum
nonwage compensation. For example, if wage by reducing on-the-job training and
workers value fringe benefits and wages flattening the tenure-wage profile (see
equally, employers can simply reduce the Jacob Mincer and Linda Leighton, 1981).
level of fringe benefits by the amount of the Indeed, one manager told our interviewer in
minimum-wage increase, leaving their em- wave 1that her workers were forgoing ordi-
nary scheduled raises because the minimum
wage was about to rise, and this would
provide a raise for all her workers. To de-
2 4 ~ i t h i nNew Jersey, the fraction of full-time em-
termine whether this phenomenon occurred
ployees increased about as quickly at stores with higher more generally, we analyzed store man-
and lower wages in wave 1. agers' responses to questions on the amountVOL. 84 NO. 4 CARD AND KRUEGER: MINIMUM WAGE AND EMPLOYMENT 787
of time before a normal wage increase and factor cost. The average restaurant in New
the usual amount of such raises. In rows 8 Jersey initially paid about half its workers
and 9 we report the average changes be- less than the new minimum wage. If wages
tween waves 1 and 2 for these two variables, rose by roughly 15 percent for these work-
as well as regression coefficients from mod- ers, and if labor's share of total costs is 30
els that include the wage-gap variable.25Al- percent, we would expect prices to rise by
though the average time to the first pay about 2.2 percent ( = 0.15 X 0.5 X 0.3) due to
raise increased by 2.5 weeks in New Jersey the minimum-wage rise.27
relative to Pennsylvania, the increase is not In each wave of our survey we asked
statistically significant. Furthermore, there managers for the prices of three standard
is only a trivial difference in the relative items: a medium soda, a small order of
change in the amount of the first pay incre- french fries, and a main course. The main
ment between New Jersey and Pennsylvania course was a basic hamburger at Burger
stores. King, Roy Rogers, and Wendy's restaurants,
Finally, we examined a related variable: and two pieces of chicken at KFC stores.
the "slope" of the wage profile, which we We define "full meal" price as the after-tax
measure by the ratio of the typical first raise price of a medium soda, a small order of
to the amount of time until the first raise is french fries, and a main course.
given. As shown in row 10, the slope of the Table 7 presents reduced-form estimates
wage profile flattened in both New Jersey of the effect of the minimum-wage increase
and Pennsylvania, with no significant rela- on prices. The dependent variable in these
tive difference between states. The change models is the change in the logarithm of the
in the slope is also uncorrelated with the price of a full meal at each store. The key
GAP variable. In summary, we can find no independent variable is either a dummy in-
indication that New Jersey employers dicating whether the store is located in New
changed either their fringe benefits or their Jersey or the proportional wage increase
wage profiles to offset the rise in the mini- required to meet the minimum wage (the
mum wage.26 GAP variable defined above).
The estimated New Jersey dummy in col-
V. Price Effects of the Minimum-Wage umn (i) shows that after-tax meal prices
Increase rose 3.2-percent faster in New Jersey than
in Pennsylvania between February and
A final issue we examine is the effect of November 1 9 9 2 . ~ T~ he effect is slightly
the minimum wage on the prices of meals at larger controlling for chain and company-
fast-food restaurants. A competitive model ownership [see column ($1. Since the
of the fast-food industry implies that an New Jersey sales tax rate fell by 1 percent-
increase in the minimum wage will lead to age point between the waves of our survey,
an increase in product prices. If we assume these estimates suggest that pretax prices
constant returns to scale in the industry, the rose 4-percent faster as a result of the
increase in price should be proportional to
the share of minimum-wage labor in total
" ~ c c o r d i n ~to the McDonald's Corporation 1991
Annual Report. payroll and benefits are 31.3 percent of
operating costs at company-owned stores. This calcula-
2 5 ~ nwave 1, the average time to a first wage in- tion is only approximate because minimum-wage work-
crease was 18.9 weeks, and the average amount of the ers make up less than half of payroll even though they
first increase was $0.21 per hour. are about half of workers, and because a rise in the
2 6 ~ a t aznd Krueger (1992) report that a significant minimum wage causes some employers to increase the
fraction of fast-food stores in Texas responded to an pay of other higher-wage workers in order to maintain
increase in the minimum wage by raising wages for relative pay differentials.
workers who were initially earning more than the new he effect is attributable to a 2.0-percent increase
minimum rate. Our results on the slope of the tenure in prices in New Jersey and a 1.0-percent decrease in
profile are consistent with their findings. prices in Pennsylvania.THE AMERICAN ECONOMIC REVIEW SEPTEMBER 1994
TABLE7-REDUCED-FORMMODELSFOR CHANGE
IN THE PRICEOF A FULLMEAL
- - - - - - - - - - - - - - - -
Dependent variable: change in the log price
of a full meal
Independent variable (i) (ii) (iii) (iv) (v)
1. New Jersey dummy 0.033 0.037 - - -
(0.014) (0.014)
2. Initial wage gapa - - 0.077 0.146 0.063
(0.075) (0.074) (0.089)
3. Controls for chain andb no yes no yes Yes
ownership
4. Controls for regionC no no no no yes
5. Standard error of regression 0.101 0.097 0.102 0.098 0.097
Notes: Standard errors are given in parentheses. Entries are estimated regression
coefficients for models fit to the change in the log price of a full meal (entrCe, medium
soda, small fries). The sample contains 315 stores with valid data on prices, wages, and
employment for waves 1 and 2. The mean and standard deviation of the dependent
variable are 0.0173 and 0.1017, respectively.
aProportional increase in starting wage necessary to raise the wage to the new
minimum-wage rate. For stores in Pennsylvania the wage gap is 0.
bThree dummy variables for chain type and whether or not the store is company-
owned are included.
'Dummy variables for two regions of New Jersey and two regions of eastern
Pennsylvania are included.
minimum-wage increase in New Jersey- One potential explanation for the latter
slightly more than the increase needed to finding is that stores in New Jersey compete
pass through the cost increase caused by the in the same product market. As a result,
minimum-wage hike. restaurants that are most affected by the
The pattern of price changes within New minimum wage are unable to increase their
Jersey is less consistent with a simple product prices faster than their competitors.
" pass-through" view of minimum-wage cost In contrast, stores in New Jersey and Penn-
increases. In fact, meal prices rose at sylvania are in separate product markets,
approximately the same rate at stores in enabling prices to rise in New Jersey rela-
New Jersey with differing levels of initial tive to Pennsylvania when overall costs rise
wages. Inspection of the estimated GAP in New Jersey. Note that this explanation
coefficients in column (v) of Table 7 con- seems to rule out the possibility that store-
firms that within regions of New Jersey, the specific demand shocks can account for the
GAP variable is statistically insignificant. anomalous rise in employment at stores in
In sum, these results provide mixed evi- New Jersey with lower initial wages.
dence that higher minimum wages result in
higher fast-food prices. The strongest evi- VI. Store Openings
dence emerges from a comparison of New
Jersey and Pennsylvania stores. The magni- An important potential effect of higher
tude of the price increase is consistent with minimum wages is to discourage the open-
predictions from a conventional model of a ing of new businesses. Although our sample
competitive industry. On the other hand, we design allows us to estimate the effect of the
find no evidence that prices rose faster minimum wage on existing restaurants in
among stores in New Jersey that were most New Jersey, we cannot address the effect of
affected by the rise in the minimum wage. the higher minimum wage on potentialVOL. 84 NO. 4 CARD AND KRUEGER: MINIMUM WAGE AND EMPLOYMENT 789
entrants.29 To assess the likely size of such tive effect on either the net number of
an effect, we used national restaurant direc- restaurants or the rate of new openings. To
tories for the McDonald's restaurant chain the contrary, all the estimates show positice
to compare the numbers of operating effects of higher minimum wages on the
restaurants and the numbers of newly number of operating or newly opened stores,
opened restaurants in different states over although many of the point estimates are
the 1986-1991 period. Many states raised insignificantly different from zero. While this
their minimum wages during this period. In evidence is limited, we conclude that the
addition, the federal minimum wage in- effects of minimum wages on fast-food store
creased in the early 1990's from $3.35 to opening rates are probably small.
$4.25, with differing effects in different states
depending on the level of wages in the VII. Broader Evidence on Employment
state. These policies create an opportunity Changes in New Jersey
to measure the impact of minimum-wage
laws on store opening rates across states. Our establishment-level analysis suggests
The results of our analysis are presented that the rise in the minimum wage in New
in Table 8. We regressed the growth rate in Jersey may have increased employment in
the number of McDonald's stores in each the fast-food industry. Is this just an anomaly
state on two alternative measures of the associated with our particular sample, or a
minimum wage in the state and a set of phenomenon unique to the fast-food indus-
other control variables (population growth try? Data from the monthly Current Popu-
and the change in the state unemployment lation Survey (CPS) allow us to compare
rate). The first minimum-wage measure is state-wide employment trends in New Jer-
the fraction of workers in the state's retail sey and the surrounding states, providing a
trade industry in 1986 whose wages fell be- check on the interpretation of our findings.
tween the existing federal minimum wage in Using monthly CPS files for 1991 and 1992,
1986 ($3.35 per hour) and the effective min- we computed employment-population rates
imum wage in the state in April 1990 (the for teenagers and adults (age 25 and older)
maximum of the federal minimum wage and for New Jersey, Pennsylvania, New York,
the state minimum wages as of April 1990)." and the entire United States. Since the New
The second is the ratio of the state's effec- Jersey minimum wage rose on April 1, 1992,
tive minimum wage in 1990 to the average we computed the employment rates for
hourly wage of retail trade workers in the April-December of both 1991 and 1992.
state in 1986. Both of these measures are The relative changes in employment in New
designed to gauge the degree of upward Jersey and the surrounding states then give
wage pressure exerted by state or federal an indication of the effect of the new law.
minimum-wage changes between 1986 and A comparison of changes in adult em-
1990. ployment rates show that the New Jersey
The results provide no evidence that labor market fared slightly worse over the
higher minimum-wage rates (relative to the 1991-1992 period than either the U.S. labor
retail-trade wages in a state) exert a nega- market as a whole or labor markets in
Pennsylvania or New York (see Card and
Krueger, 1993 table 9l3' Among teenagers,
29
however, the situation was reversed. In New
Direct inquiries to the chains in our sample re- Jersey, teenage employment rates fell by 0.7
vealed that Wendy's opened two stores in New Jersey
in 1992 and one store in Pennsylvania. The other percent from 1991 to 1992. In New York,
chains were unwilling to provide information on new
openings.
30
We used the 1986 Current Population Survey
31
(merged monthly file) to construct the minimum-wage The employment rate of individuals age 25 and
variables. State minimum-wage rates in 1990 were ob- older fell by 2.6 percent in New Jersey between 1991
tained from the Bureau of National Affairs Labor and 1992, while it rose by 0.3 percent in Pennsylvania,
Relations Reporter Wages and Hours Manual (undated). and fell by 0.2 percent in the United States as a whole.You can also read
